l,4,8,1 l]tetraazacyclotetradecine Macrocyclic Ligand - American

Comparative Electrochemical. Behavior of the [ M d M o ] Core. J. M. KerbaoI,l1 E. Furet,lb J. E. Guerchais,*J1 Y. Le Mest," J. Y. SaiUard,*Jb J. Sala...
0 downloads 0 Views 1MB Size
713

Inorg. Chem. 1993, 32, 713-721

Bridging Bis-Bidentate Coordination of the Usually Tetradentate 5,7,12,14-Tetramethyldibenz~b,~l,4,8,1 l]tetraazacyclotetradecine Macrocyclic Ligand: Experimental and Theoretical Study of [ M 0 2 ( 0 2 C R ) 2 ( C 2 2 H 2 2 N 4 ) ] . Comparative Electrochemical Behavior of the [ M d M o ] Core J. M. KerbaoI,l1 E. Furet,lb J. E. Guerchais,*J1Y. Le Mest," J. Y. SaiUard,*JbJ. Sala-Pala,'' and L. ToupetlC Unit€ AssociQ au CNRS 322, Laboratoire de Chimie, Electrochimie Molkulaires et Chimie Analytique, Universite de Bretagne Occidentale, 29287 Brest Cedex, France, Unit€ AssociQ au CNRS 254, Laboratoire de Chimie du Solide et Inorganique MolQulaire, UniversitC de Rennes I, 35042 Rennes Cedex, France, and Unit6 Assmike au CNRS 804, Groupe Mati8re CondensQ et Materiaux, Universit6 de Rennes I, 35042 Rennes Cedex, France

Received March 18. 1992 Treatment of the dimolybdenum tetracarboxylato derivatives [M02(02CR)4] (2)(2a, R = CH3; 2b,R = C(CH3)3) with the H2TMTAA (C22H24N4) ligand in pentan-1-01 in the presence of DBU affords the new complexes [Moz(TMTAA)(02CR)2] (1) (la, R = CH3; lb, R = C(CH3)3). Complex la has been structurally characterized. Crystal data for [MO~(C~ZH~~N~)(O~CCH~)~]: triclinic, space group Pi,u = 8.801(5) A, b = 11.572(8) A, c = 13.714(9) A, a = 108.42(6)O, B = 90.29(5)O, 7 = 106.35(5)O, Y = 1265(1) A3, Z = 2, Dc = 1.71 g-cm-3, F(Oo0) = 824, p(Mo Ka) = 10.06 cm-I, and R = 0.039 ( R , = 0.034)for 410 parameters and 2973 unique data with I > u(I). The structure of la consists of discrete dinuclear molecules in which the macrocyclic ligand has a very unusual bridging bis-bidentate coordination while the two acetato groups act as bridging ligands as in 2. The two [MoOZN~] moieties are in a quasi-perfectly eclipsed geometry, and the metal-metal bond length [2.106(1) A] is in agreement with a uz#tS2 quadruple bond. SCF-MS-X, calculations performed on a simplified C, model show that the G(MoMo) orbital (HOMO) is definitely bonding and the S*(MoMo) orbital (LUMO) definitely antibonding with a significant HOMO/LUMO gap of 0.90 eV. The electrochemicalbehavior of complexes 1was investigated. They are oxidizable in two one-electron reversible steps, leading successively to a stable mixed-valence [Mo"%Mo'I']+ compound and to a [Mo"'~Mo"']~+species which decomposes readily. Two successive oneelectron reduction steps lead to unstable [Mo"=MoI]- and [Mo'-?-Mo']'- derivatives. The electrochemistryof the [(Mo(TMTAA)]2] derivative 3 has been reexamined and compared to that of compounds 1 and 2. The effect of the coordination mode of the macrocycle ligand on the electrochemicalreactivity of the [ M d M o ] core is discussed.

Introduction Although superficially resembling a porphyrin, the 5,7,12,14tetramethyldibenzo[b,i][ 1,4,8,11]tetraazacyclotetradecinemacrocyclic ligand, HzTMTAA (Cz2H24N4; Figure l), shows significant differences in electronic delocalization, core size, and framework flexibility.24 These structural differences clearly induce differencesin reactivity. This is well illustrated by their reaction with molybdenum hexacarbonyl: while H2TMTAA (H2L) affords a tetracarbonyl complex [Mo(C0)4(H2L')] in which the N4 macrocycle adopts very unusual behavior (neutral bidentate coordination with H transfer from a NH moiety to a CH unit),S-' H2TPP gives rise to the quadruply metal-metal bonded molybdenum(I1) complex [(Mo(TPP)]2] in which the porphyrin adopts its classical tetradentate dianionic form.* However, the TMTAA analog of this complex, Le. [(Mo(TMTAA)]2] (3) (Figure 2), has been obtained by reaction of (1) (a) URA CNRS 322. (b) URA CNRS 254. (c) URA CNRS 804.

(2) (a) Jager,E.G.Z.Anorg.Allg.Chem.1969,361,177. (b)L'Epplatenier, F. A.; Pugin, A. Helu. Chim. Actu 1975.58.917. (c) Goedken, V. L.; Weiss, M. C. Inorg. Synth. 1980, 20, 115. (3) Goedkcn, V.L.; Pluth, J. J.; Peng, S. M.; Bursten, B. J . Am. Chem. Soc. 1976, 98,8014. (4) For a recent review on m o m and dinuclearcomplexesofthe H2TMTAA ligand, see: Cotton, F. A.; Czuchajowska, J. Polyhedron 1990,9,2553. (5) Bell, L.G.;Dabrowiak,J.C. J . Chem.Soc..Chem.Commun. 1975,512. ( 6 ) Giraudon, J. M.; Mandon, D.; Sala-Pala, J.; Guerchais, J. E.; Kerbaol, J. M.; Le Mest, Y.; L'Haridon, P. Inorg. Chem. 1990, 29,707. (7) Cotton, F. A,; Czuchajowska, J. Polyhedron 1990, 9, 1217. (8) Yang, C. H.; Dzugan, S.J.; Goedken, V. L. J . Chem. Soc., Chem. Commun. 1986. 1313.

H 3C"CH3

Figure 1. Comparison of Nd macrocyclic ligands: Left, the 5,7,12,14tetramethyldibenzo[b,i] [ 1,4,8,1 I] tetraazacyclotetradcine macrocyclic ligand, HzTMTAA (C22H24N4); right, the porphyrin ligand. For H2TPP, R = C ~ H and S R' = H.

i

L

Figure 2. Syntheses of the molybdenum complexes from [Mo2(02CR)4] (2). Key: (i) HzTMTAA (H2TMTAA/2 = 2), pentan-1-01, reflux in presence of DBU gives [Mo~(O~CR)~(TMTAA)] (1) (yield ca. 50%) (this work); (ii) Li2TMTAA (Li2TMTAA/2 = 2),THF, -30 OC, gives [(Mo(TMTAA))2] (3) (yield ca. 70%) (see ref 9).

dimolybdenum tetraacetate, [ M02(02CCH3)4], with the lithium salt Li2TMTAA and fully characteri~ed.~J~ (9) Mandon, D.;Giraudon, J. M.; Toupet, L.; Sala-Pala, J.; Gucrchais, J. E.J . Am. Chem. Soc. 1987,109, 3490. (10) Cotton, F. A.; Czuchajowska, J.; Feng, X. Inorg. Chem. 1990,29,4329.

0020-1669/93/1332-0713$04.00/0 Q 1993 American Chemical Society

714 Inorganic Chemistry, Vol. 32, No. 5, 1993

Kerbaol et al.

In a continuation of our investigations on the ability of the H2T M T A A ligand to bind molybdenum ions, we have scrutinized

the reaction between [Mo2(02CCH3)4] and the free macrocycle H2TMTAA. For comparison, we have also examined the reaction of H2TMTAA with [Mo2(02CC(CH3)3)4], in which the very highly electron-donating tert-butyl group C(CH3)3 is present. As the analytical data for the resulting molybdenum(11) complexes 1 clearly showed an unusual TMTAA/metal ratio of 0.5, the solid-state molecular structure of the acetato complex was determined, to clearly establish the coordination mode of the N4 ligand. Subsequently,this complex was further investigated by theoretical studies. The electrochemical behavior of these new compounds 1 was also investigated. The electrochemistry of the previously described [{Mo(TMTAA)J2]derivative 3, shown in Figure 2, was reexamined more thoroughly in the present work in addition, since some misinterpretationwas found in the previous work.9 The electrochemistry of the tetracarboxylato derivatives [ M O ~ ( O ~ C R ) and ~ ] Iof ~ Jother ~ [ M o ~derivatives ] with different ligands” has been previously described; the properties of the two [Mo2(02CR)4] complexeswith R = CH3 and R = C(CH3)3 have been reexaminedin details in the present experimental conditions. Together with the well-known tetracarboxylato derivatives [MO~(O~CR , the ) ~ ] [Mo~(TMTAA)(O~CR)~] and [(Mo(TMTAA))z] derivatives constitute an interesting series to examine the effect of the coordination mode of the macrocyclicligand on the electrochemical reactivityof the [ M d M o ] core; the results are discussed in the light of the MO diagram for these three kinds of compounds. Although some porphyrin complexesof the [M d M o ] core*J4are known, their electrochemical properties have not yet been described; the ‘Mo2(TMTAA),” complexes are thus the first examples of macrocyclic derivatives of the [ M d M o ] core whose electrochemistryis described.

Experimental Section All reactions were performed in Schlenk tubes in a dry oxygen-free nitrogen atmosphere. Solvents were distilled by standard techniques and thoroughly deoxygenated before use. Elemental analyscs were performed by ‘Service Central d’Analyse du CNRS-Vernaison, France”. IR spectra were recorded as KBr and CsI pellets with a Perkin-Elmer 1430 spectrophotometer. ESR spectra were obtained using a Jeol FE 3X spectrometer (X-band). The apparatus, the three-compartment microelectrochemical cell, the treatment of the solvent, and the supporting electrolyte for the electrochemical studies have all been previously described.I5J6 The potentials are quoted versus the ferrocenium/ferrocene redox (Fc+/Fc) system. For comparison of the present results with other works, the Fc+/Fc formal potential has been measured versus SCE in bcnzonitrile (0.2 M, Bu4NPF6), E’O(Fc+/Fc) +0.43 V vs SCE. ~

~

~

~

~

~

~

~~~~~

~~

(1 1) For a review, see: Casewitt, C. J.; Rakowski-Dubois, M. Rev. Inorg. Chem. 1988, 9, 199. (12) (a) Cotton, F. A.; Pedersen, E. Inorg. Chem. 1975,14,399. (b) Santure, D. J.; Huffman, J. C.; Sattelkrger, A. P. Inorg. Chem. 1985, 24, 371. (c) Cayton, R. H.; Chisholm, M. H.; Huffman, J. C.; Lobkovsky, E. B. J. Am. Chem. Soc. 1991, 113, 8709. (13) (a) Cotton, F. A.; Frenz, B. A.;Pedersen, E.;Webb,T. R. Inorg. Chem. 1975,14,391. (b) Zietlow, T. C.; Hopkins, M. D.; Gray, H. B. J. Am. Chem. Soc. 1986. 108, 8266. (c) Zietlow, T. C.; Klendworth, D. D.; Nimry, T.; Salmon, D. J.; Walton, R. A. Inorg. Chem. 1981, 20, 947. (d) Coffindaffer, T. W.; Niccolai, G. P.; Powell, P.; Rothwell, I. P.; Huffman, J. C. J. Am. Chem. Soc. 1985, 107, 3572. (14) (a) Tai, C. D.; Garner, J. M.; Collman, J. P.; Sattelberger, A. P.; Woodruff. W. H. J . Am. Chem. Soc. 1989. I l l . 9072. (b) Collman. J. P.; Kim, K.; Garner, J. M. J. Chem. So;., Chem. Commun. 1986; 1711. (c) Collman, J. P.; Woo, L. K. Proc. Natl. Acad. Sci. U.S.A. 1984. 81. 2592. (15) Le Mest,’Y.; L‘Her, M.; Courtot-Coupez,J.; Collman, J. P.; Evitt, E. R.; Bencosme, C. S.J. Electroad. Chem. Interfacial Electrochem. 1985, 184, 331. (16) Giraudon, J. M.; Sala-Pala, J.; Guerchais, J. E.; Le Mest, Y . ;Toupet, L. Inorg. Chem. 1991, 30, 891; Err. 1991, 30, 3112.

Table I. CrvstaiionraDhic Data for la chem formula C26H2nM%N404 fw 652.4 space group Pi a, A. 8.801(5) b, A. 11.572(8) c, A0 13.714(9) a,deg’ 108.42(6) 90.29(5) 8, dego 106.35( 5) 7,deg’

~

v,A3 Z T , OC

LA

p C p l ~gem-’ .

k,cm-1

Rb RwC

~

~~

1265( 1) 2 23 0.710 69 1.71 10.06 0.039 0.034

- From a least-squares fitting of the setting angle of 25 reflections. R - xlFo - F#xF0. R , = [E@, - Fc12/EwFo2]1/2,with w = (d(I)+ a

(0.04F02)2)-1/2.

H2TMTAA was prepared as previously described by Goedken and Weiss.2c Complexes [Mo2(02CR)4] (2) were prepared according to the procedure of Stephenson et al. (R = CHI) and McCarley et ai. (R = C(CH3)d.‘’ Complex la. A mixture of [Mo2(02CCH3)4] (2a) (428 mg, 1 mmol) and H2TMTAA (688 mg, 2 mmol) in pentan-1-01 (60 mL) was refluxed under dinitrogen for ca. 7 h in the presence of DBU (1,8-diazabicyclo[ 5.4.0lundec-7-ene; C9H16N2; 2 mL). The resulting solution was then slowly cooled to room temperature. The violet solid l a was filtered out, washed with pentane, and dried under r e d u d pressure (yield ca. 50%). Anal. Found: C, 47.1; H, 4.3; Mo, 29.9; N, 8.3. Calcd for C26H28Mo~N404:C, 47.9; H, 4.3; Mo, 29.5; N, 8.6. IR data (cm-I): 296 m, 335 s, 515 m, 625 m 6(OCO), 677 s 6(OCO), 745 s, 788 s, 832 m, 930 w, 1018 s, 1040 w, 1102 w, 1195 s, 1235 m, 1270 m, 1350-1550 br. Complex lb. l b was prepared from 2b by a procedure similar to that described above for l a (yield ca. 50%). Anal. Found: C, 53.5; H, 5.7; Mo, 25.5; N, 7.8. Calcd for C32HmMo2N404: C, 52.2; H, 5.4; Mo, 26.1 ; N, 7.6. IR data (cm-I): 425 m, 610 m, 690 w, 740 s, 755 w, 780 s, 830 m,895m,930m,1015m,1100w,1150w,1190s,1220s, 1270m,1355 s, 1370 s, 1410 vs, 1468 vs, 1482 s, 1504 s, 1545 vs. Determination of tbe Crystal Structure of la. Data CoUectioa A violet crystal, ca. 0.20 X 0.15 X 0.20 mm, was sealed under dinitrogen in a capillary and placed on an Enraf-Nonius CAD-4 diffractometer, with graphite-monochromated Mo K a radiation. All crystallographic data are given in Table I. Cell parameters were determined by least squares on setting angles from 25 reflections. Data were collected (w-28 scan; tmal = 60 s) in the following hkl range: h, 0 to 10; k, -13 to 13; I, -16 to 16 ( 2 ~ 9 , =~ ~55’). Intensity variation over the period of data collection: i0.4%. A total of 4456 reflections were measured, 2973 unique with I > u(I) (Rint= 0.016). A Lorentz polarization correction was performed. Determination and Refmment of the Structure. The structure was solved by Patterson and Fourier methods and refined by full-matrix least squares using anisotropic thermal factors for all non-hydrogen atoms while hydrogen atoms were refined isotropically. Refinement (2973 data, 410 parameters) was terminated with R = 0.039 and Rw = 0.034 [w = ( 0 2 ( r )+ (0.04F02)2)-1/2].The final difference map was flat [IAdmaX = 0.32 PA-31. Atomic scattering factors were taken from ref 18. Calculations were performed on a Digital PDP 11/60 computer by using the SDP package of crystallographic program.Ig Final atomic parameters are listed in Table 11, and interatomic bond distances and angles in Table 111. Molecular Orbital Cdculrtioas. All the calculations have been made within the standard SCF-MS-X, method.20 The model M chosen for the calculations is a simplification of la, in which the acctato ligands have been replaced by formato groups and TMTAA was replaced by the hypothetical CloHloN4ligand, using the same procedure as Cotton et al. This model is shown in their theoretical study of [(MO(TMTAA))~].~O

.

(17) (a) Stephenson, T. A,; Bannister, E.; Wilkinson, G. J. Chem. Soc. 1964, 2538. (b) McCarley, R. E.; Templeton, J. L.; Colburn, T. J.; Katovic, V.; Hoxmeier, R. J. Adv. Chem. Ser. 1976, No. 150, 318. (1 8) International Tablesfor X-ray Crystallography; Kynoch: Birmingham, England, 1974; Vol. IV. (19) (a) Frenz, B. A. ENRAF-NONIUSStructure Determination Package; SDP Users guide, Version 1985. College Station, TX 77840 and ENRAF-NONIUS, Delft, The Netherlands. (b) Johnson, C. K. ORTEP. Report ORNL-3794, Oak Ridge National Laboratory, Oak Ridge, TN. (20) (a) Slater, J. C. Adv. Quantum Chem. 1972.6, 1. (b) Johnson, K.H. Adv. Quantum Chem. 1973,7,143. (c) Slater, J. C. Quantum Theory of Molecules and Solids; McGraw-Hill: New York, 1974; Vol. 4. (d) Johnson, K. H. Annu. Rev. Phys. Chem. 1975, 26, 39.

Inorganic Chemistry, Vol. 32, No. 5, 1993 715

[MOZ(~~CR)Z(C~ZH~ZN~)J Tabk 11. Final Fractional Atomic Coordinates in la atom

X

Y

Z

Mol Mo2 01 02 03 04 N1 N2 N3 N4

-0.09759(51 -0,.1003sisj -0.3440(4) -0).3479(4) -0.1509(4) -0).1533(4) -0.0366( 5 ) -0.0409(5) 0.15 14(5 ) 0.1478(5) 0.4144(7) 0.2445(6) 0.1971 (6) 0.0796(6) 0.0965(7) 41378(6) -0.1372(6) -0,241 l(7) -0.3405(7) -0.3429(7) -0.2430(7) -0.4 167(6) -0.5879(8) 0.4095( 7) 0.2397(6) 0.1918(6) 0.0739(6) 0.0921(7) 0.21 lO(6) 0.2579(8) 0.3011(8) 0.2952(8) 0.2476(7) 0.2062(6) -0.1634(6) 4.187( 1)

0.02454(41 0.01406('4j -0.0736(3) -0.0860(3) 0.2005(3) 0.1897(3) -0.1464(3) -0.1601 (4) 0.1223(3) 0.1067(4) 0.1573(5) 0.079 1( 5 ) -0.0441 ( 5 ) -0.15 17(4) -0.2784( 5 ) -0.2576(4) -0.2663(4) -0.3727(5) -0.4697(5) -0.4600(5) -0.3530(5) -0.1 -0.1864(7) 0.1346(6) 0.0609(5 ) -0.0656(5) 4.1710(4) -0.3004(5) 0.2383(4) 0.3572(5) 0.467 1( 5 ) 0.4594(5) 0.3421(5) 0.2292(4) 0.2482(5) 0.3750(6)

0.83188(31 0.67569(3j 0.8120(3) 0.6471(3) 0.8765(3) 0.7124(3) 0.8 157(3) 0.6096(3) 0.8814(3) 0.675 l(3) 0.9673(5) 0.9271(4) 0.9369(4) 0.8738(3) 0.8715(4) 0.7402(4) 0.6357(4) 0.5623(4) 0.5905(5) 0.6925(5) 0.7676(4) 0.7229(4) 0.7042(5) 0.6078(5) 0.6088(4) 0.5386(4) 0.5482(4) 0.4872(5) 0.8574(4) 0.9320(4) 0.9064(5) 0.8060(5 ) 0.7288(4) 0.7532(4) 0.8063(4) 0.8333(5)

c1 c2 c3 c4 c5 C6

c7 C8 c9 c10 c11 c12 C13 C14 C15 C16 C17 C18 C19 c20 c21 c22 C23 C24 C25 C26

E,' A2 1.92(1) 1.92(1) 2.65(8) 2.64(8) 2.77(8) 2.74(8) 2.07(9) 2.26(9) 2.13(9) 2.22(9) 3.8(1) 2.4(1) 2.4(1) 2.1(1) 3.1(1) 2.3(1) 2.3(1) 3.1(1) 3.5( 1) 3.4(1) 2.9(1) 2.7(1) 5.0(2) 3.8(2) 2.5(1) 2.7(1) 2.3(1) 3.4(1) 2.5(1) 3.4( 1) 4.1(2) 3.9(1) 3.2(1) 2.4( 1) 2.9( 1) 5.5(2)

Table 111. Coordination S here of the Molybdenum Atoms in la: Selected Bond Distances and Bond Angles (deg)' Mo-Mo2 2.106( 1) m0242 2.127(4) Molal 2.116( 3) Mo2-04 2.1 19(4) Mol43 2.120(3) M02-N2 2.145 (4) Mol-NI 2.139(4) MO2-N4 2.146(4) MOl-N3 2.156(4)

(1)

Mo2-Mol41 MO2-Mol43 M02-Mol-Nl M02-Mol-N3 01-Mol43 01-Mol-N 1 Ol-Mol-N3 03-Mol-N 1 03-MOl-N3 Nl-Mol-N3

91.4( 1) 92.0(1) 98.1(1) 99.4(1) 89.6( 1) 91.8(2) 169.1( 1) 169.8( 1) 90.9(2) 8 5.8( 2)

Mol-M02-02 Mol-MO2-04 Mol-MoZ-NZ Mol-M02-N4 02-M02-04 02-M02-N2 02-M02-N4 04-M02-N2 O&Mo2-N4 N2-Mo2-N4

91.62(9) 90.93(9) 100.0(1) 9 8 4I) 90.0( 1) 91.3(2) 169.8( 1) 168.9( 1) 91.5(2) 8 5.3(2)

Estimated standard deviations in the last significant figure are given in parentheses.

h

E values for anisotropically refined atoms are given in the form of the ieotropicquivalent displacementparameter defined as (4/3) [&E( 1,l) + b2B(2,2) + c2E(3,3) + ob(- y)E(1,2) + ac(cos P)E(1,3) + bc(cos

A

a

a)B(2,3)1. in Figure 3. The structural parameters of M were determined using the following procedure: Starting from the structure of la determined by X-ray diffraction, the crystallographic coordinates were converted into a set of Cartesian coordinates. The origin of this Cartesian system was chosen to be the middle of the MoMo bond; the z axis was chosen to be colinear to the MoMo bond, and the x axis was chosen to be normal to a plane defined as the average between the least-squares planes of the four N atoms and the four 0 atoms. At this stage, the idealized Cb symmetry was introduced by a coordinate averaging of the symmetryrelated atoms. It was checked that the bond distances and angles of the model structure M are not significantly different from the corresponding averaged experimental values of la (largest difference in bond distances 0.0014 A). All the C-H bond distances were set to 1.09 A. The muffin-tin atomic spheres were allowed to overlap. Due to the geometry, different reduction factors values for the atomic radii had to be considered,trying to keep an quilibrium &tween the electroniccharge of the atomic spheres and their overlap, especially for the carbon atoms of the simplified TMTAA ligand. We used 90% of the atomic number radii for the core of the molecule (is. for Mo, 0, N) and 82.5% for C. The hydrogen radius was fixed to 0.87 bohr, a value that has been computationally tested by us on several molecules. As usual, the outer sphere was taken close to the outermost atomic sphere, and its center was placed out of the MoMo axis, leading to the smallest possible intersphere region. The sphere radii are given in Table IV. The a parameters used (Table IV) are the Q H F of Schwarz,2' except for hydrogen, where the Slater value22was chosen. Those relative to the outer-sphere (extramoIecular) and the intersphere regions are weighted-average values of the atomic ones. Potential waves up to I = 4 were used for the outer sphere, up to I = 2 for the metal atoms, and up to I = 1 for 0, C, and N. Only the I = 0 value was considered for H. (21) (a) Schwarz, K.Phys. Reu. E 1972.5,2466. (b) Schwarz, K. Theorer. Chim. Acta 1974,34, 225. (22) Slater, J. C. Inr. J . Quunrum Chem. 1973, 7, 533.

Figure 3. View of the model M used in the SCF-MS-X, calculations. The atom labels refer to Table IV. Table IV. SCF-MS-X, atom tYW' intersphere outer sphere Mo 0 N a

Parameters Used in the Calculations radius

a

0.759 73 0.759 73 0.703 41 0.74447 0.751 97

bo) 9.3325 2.3426 1.7567 1.7217

atom type' CI Cz C, Cq

H

radius a

(00)

0.759 28 0.759 28 0.759 28 0.759 28 0.777 25

1.6259 1.4733 1.5776 1.6159 0.8700

See Figure 3.

Results and Discmion Syntksis and IR Data. Treatment of dimolybdenum tetraacetate [Mo2(02CCH3)4] (h) with 2 equiv of the 5,7,12,14tetramethyldibenzo[1,4,8,11J tetraazacyclotetradecineligand HzTMTAA ( C Z ~ H Z ~Figure N ~ ; 1) in pentan-1-01 at reflux in the presenceof DBU (1,8-diazabicyclo[5.4.0]undec-7-ene;C ~ H I ~ N Z ) caused precipitation, after slow cooling to room temperature, of a violet solid l a in ca. 50% yield. IR spectroscopyclearly showed differences between l a and the previously reported dinuclear complex [(Mo(TMTAA))z] (3), which we obtained from reaction of 2a with the lithium derivative [L~z(TMTAA)J.~ A complete interpretation of the vibrational spectrum of l a was very difficult since superposition of various strong absorptions occurred in the 1350-1600-~m-~ region, giving a poorly resolved broad signal. If the presence of the macrocyclic ligand in l a was deduced from sharp characteristic absorptions at 745 and 1195 cm-I, that of the acetato groups did not appear very clearly despite a careful comparison with the IR spectrumof the starting dinuclear complex

Kerbaol et al.

716 Inorganic Chemistry, Vol. 32, No. 5, 1993

Table V. Comparison of Average Dihedral Angleso (deg) and Selected Average Interatomic Distances (A) in (Mo(TMTAA)] Complexesb bis-bident TMTAA tetradent TMTAA 1-2 angle(av)' 1-3 angle(av)' 1 4 angle(av)O dist of Mo from N4 plane Mo-N dist from to av

le

Id

48.9(2) 9.2(8) 13.6(4) 1.540(1)

32.2 13.9 21.4 0.570(5)

2.139(4) 2.156(4) 2.146(4)

2.057(4) 2.079(4) 2.068(4)

5c

33.8 22.5 0.737(1) 2.034(4) 2.041(4) 2.038(4)

Osee definition of dihedral angles in Figure 5. bSimilar data for TMTAA complexes with other metals are given in ref 29. 1, [MO~(O~CCH~)~(TMTA this A work. )]; 4, [(Mo(TMTAA))2]PF6;see ref 9. 5, [MoO(TMTAA)][Cs(CN)s]; see ref 16. U

Figure 4. View of the structure of complex [ M o ~ ( T M T A A ) ( O ~ C C H ~ ) ~ ] (la), showing the atom-labeling scheme. Hydrogen atoms are omitted for clarity. Non-hydrogen atoms are shown as 50% thermal ellipsoids.

[Mo2(02CCH3)4] (2a).23However, the presence of 02CCH3 groups, suggested by the absorption at 677 cm-I attributable to a(OC0) ~ibration,~3 was corroborated by elemental analyses which were in agreement with the formulation of la as [Moz(TMTAA)(02CR)z] (R = CH3). The analog lb with the highly electron-donating R = C(CH& group was similarly obtained by reacting H2TMTAA with the tetrapivalato complex [MO~(OZCR)~] (Zb,R = C(CH3M. Since H2TMTAA usually gives 1/1 metal complexe~,~ the metal/macrocyclicligand ratioobserved in complexes 1appeared extremely unusual, even though a similar value of two had been previously r e ~ a r t e d ? ~for , ~example ~ in rhodium chemistry for the [Rh2(TMTAA)(CO)4] and [Rhz(HTMTAA)(C0)4]C104 complexes.24 It is noteworthy (i) that complexes 1were prepared from reactions in which the starting Mo/HzTMTAA ratio was one and not two and (ii) that all attempts to obtain [(Mo(TMTAA))2] from reaction of [ M o ~ ( O ~ C R (2) ) ~ ] with H2TMTAA failed, even when using a large excess of ligand, while displacement of all acetato ligands from h easily occurred upon reaction with L~zTMTAA.~ Structural Results. Figure 4 shows a perspective view of the molecule l a with the atomic numbering scheme. The structure consists of discrete dinuclear complex molecules with the macrocyclic ligand in a bridging bis-bidentatecoordination mode, with each Mo(I1) bonded to one of the 2,4-pentanediiminato chelate rings. The coordinationaround each molybdenum atom is completed by two bridging acetato groups cis to each other, the two M002N2 fragments being in a quasi-perfectly eclipsed geometry. The maximum point symmetry permitted with this ligand arrangementis Ck,and the dinuclear molecule has virtually this symmetry. Therefore, complex la has a structure which shows important analogies with that of the starting dinuclear acetato complex b. For example, from a structural point of view, formation of 1 from 2 may formally appear as a simple substitutionof two bridging carboxylato groups by the bis-bridging dianionicTMTAA ligand without significant structural changes Clark, R. J. H.; Hempleman, A. J.; Kurmoo, M. J . Chem. Soc., Dalron Tram. 1988, 973 and references therein. Gordon, G.C.; DeHaven, P. W.; Weiss, M. C.; Goedken, V. L. J . Am. Chem. Soc. 1978, 100, 1003. Other examples of H2TMTAA complexes having a M/TMTAA ratio equal to two have been also reported. (a) [(Re(CO)3)2(TMTAA))has been briefly described by: Kadish, K. M.; Bottomley, L.A.; Schaeper, D.; Tsutsui, M.; Bobsein, R. L.Inorg. Chim. Acra 1979, 36, 219. (b) [Ru2(HTMTAA)(PhCCPh)(02CCH3),] has been obtained by: Cotton, F. A.; Czuchajowska, J. (J. Am. Chem. Soc. 1991, 113, 3427) from reaction of [ R u ~ ( O ~ C C Hwith ~)~ H2TMTAA ] in presence of PhCCPh. Actually, the crystal structure study of this complex revealed a coupling reaction between the HTMTAA and PhCCPh fragments.

Figure 5. Definition of molecular planes. The dihedral angles for la and other [Mo(TMTAA)] complexes are given in Table V.

and particularly without significant modification in the metalmetal bond length. As in 2 (MoMo = 2.0934(8) A),26 the metalmetal bond length in l a (2.106(1) A) is in agreement with a a 2 d 6 2quadruple bond.27 The structure of complex la may be also roughly compared to that of the previously described [Rh2(TMTAA)(CO),] complex, with the two bridging acetato groups taking the place of the four terminal carbonyl ligands.% However, and as expected, these two complexes strongly differ in the nature of the metalmetal interaction since only a weak bonding interaction (RhRh = 3.086 A) was observed in the rhodium deri~ative.2~ The macrocyclic ligand displays a geometry which corresponds to a significantdeformationfrom its usually stable saddleshape.3~~ The four nitrogen atoms are coplanar within 0.004 A, and the Moz fragment is parallel to this plane (Mo-N4 mean plane distances: 1.539(1)and 1.541(1) A). It isnoteworthy that these displacements are much greater than those observed in [(Mo(TMTAA))2]PF6(4) (0.570(5) A)9 and [MoO(TMTAA)] [Cj(CN)s] (5) (0.737( 1) A),16in which themacrocyclicligand shows its usual tetradentate coordinationmode (Table V). The average Mo-N bond length in l a (2.146(4) A) is considerably longer than the averages of 2.068(4) and 2.038(4) A found in 4 and 5, respectively. It is of prime interest to note that in going from the tetradentate to the bis-bidentate coordination mode, the most significantchange in the geometry of the ligand corresponds to a significant twisting about the C-N bonds of the 2,4-pentanediiminatochelate rings, the angle between these two rings increasing from ca. 65O to ca. 98' (Table V). On the other hand, the average angle between the two benzene rings decreases from ca. 44O in 4 and 5 to ca. 27O in 1. As in other complexes of the TMTAA ligand,3,4v9J6the four CN bonds of the @phenylenediamineresidues (average length 1.43l(6) A) have predominant single-bond character and break the electronic delocalization that is primarly limited to the benzene rings and the NCCCN units. (26) Cotton, F. A.; Mester, Z. C.; Webb, T. R. Acta Crysrallogr., Sect. B 1974,830. 2168. (27) (a)Cotton,F.A.;Curtis,N.F.;Harris,C.B.;Johnson,B.F.G.;Lippard, S. J.; Mague, J. T.; Robinson, W. R.; Wood, J. S.Science 1964, 145, 1305. (b) Cotton, F. A. Inorg. Chem. 1965.4, 334.

Inorganic Chemistry, Vol. 32, No. 5, 1993 Table VI. Valence Molecular Orbital Data for the Cb Model

M of l a

% charge

levels

E, eV -3.296 -3.677 -3.812 -3.934 -4.541 -4.700 -4.988 -5.233 -6.371 -7.270 -7.927 -8.341 -8.552 -9.208 -9.424 -9.523 -9.678 -10.1 1 -10.13 -10.44 -10.51 -10.75

int. 69 47 59 20 39 26 17 22 11 19 25 30 27 31 20 17 17 17 23 16 17 13

out. 10 6 6 1 2 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0

717

M02 19 32 5 35 2 27 76 53 75 48 32 5 14 12 32 54 62 1 17 3 0 53

The [Mo2(02CCH3)2]fragment reveals nounexpected features and contains almost perpendicular acetato groups. It is noteworthy that the Mo-0 bond distances, ranging from 2.226(3) to 2.127(4) A, are comparable to those observed in other [Mo& acetato),] complexes.z* Theoretical Analysis. SCF-MS-X, calculations have been performed on a simplified CZ, model of complex 1, namely M, shown in Figure 3. The way the structure of this complex has been set up, and the calculationshave been carried out, is discussed in the Experimental Section. The energy, symmetry, and charge distribution of the highest occupied and lowest unoccupied molecular levels are given in Table VI. Under the CZ, symmetry of the molecule, the u (d12) and 6 (dX2+) metallic orbitals belong to the same a l irreducible representation, as does one component of the T orbitals (dJ that we shall name T,. Therefore, these 6, u, and T , orbitals are allowed to mix in the complex. However, this mixing is not important, due to the strong +fold pseudosymmetry around the metal atoms. Although the metallic levels can also mix with ligand levels of the same symmetry, which lie close in energy, it is possible to identify unambiguously the bonding and antibonding orbitals which can be associated with the u, 6, A,, and r MoMo bonds (Table VI). This is illustrated in Figure 6, which shows plots of all the metal-metal bonding combinations and of the 6* antibondingorbital. Fromthesedatait appearsthat theG(MoMo) orbital (which is the HOMO of M) is definitely bonding and that the G*(MoMo) orbital (which is the LUMO) is definitely antibonding. In accord, the HOMO/LUMO gap is significant (0.90 eV). It is clear that a quadruple bond is present in this molecule, a finding which agrees with the observation of a short MoMo distance in la (2.106(1) A). It is interesting to compare our theoretical results with those obtained by the same method of calculation on related complexes containing MoMo quadruple bonds.30 Cotton et aLiOhave carried out X, calculations on a model for [(Mo(TMTAA)J2] (3), in which each Mo atom is bonded to the four nitrogen atoms of the same TMTAA ligand. Although the occupied MoMo 6 orbital (28) Cotton, F. A.; Isley, W. H.; Kaim, W. Inorg. Chem. 1981,20,930and references therein. (29) Tsutsui, M.; Bobsein, R. L.; Cash, G.; Pettersen, R. Inorg. Chem. 1979, 18, 758 and references therein. (30) For a rccent and general discussion on the theoretical aspects of the metal-metal quadruple bond, see: Wiest, R.; Strich, A.; Benard. M. New J . Chem. 1991, 15, 801 and references therein. See also: Losada, J.; Alvarez, S.;Novoa, J. J.; Mota, F.;Hoffmann, R.;Silvestre, J. J . Am. Chem. Soe. 1990,112,8998.

Mo angular contribn (acet)Z 1 8 0 5 0 0 1 2 6 3 1 2 6 3 40 15 14 42 36 71 75 14

tmtaa 1 8 30 39 57 46 6 23 7 30 41 63 53 53 7 14 6 39 23 10 7 20

%S

u*(Mo-Mo)

%D

%d

11

89 100 100

1 1

r,*(Mo-Mo) r*(Mo-Mo) B*(Mo-Mo) B(Mo-Mo)

r(Mo-Mo) r,(Mo-Mo)

u( Mo-Mo)

99 99 100 100 100 98

1

1

18 12

7 3 3

2

3

93 79 85 100 95

9

91

4

90

6

of this complex appears significantly bonding, the HOMO/ LUMO gap (0.38 eV) is about half the one we found on 1.This result suggests that the 6 bond in 1 is stronger than that in 3, consistent with the fact that the MoMo bond distance in [(Mo(TMTAA)]2] is longeri0 (2.22 A vs 2.1 l A). Norman et ala3'have carried out X, calculations on [Moz(OZCH)4], in which each formato ligand is bonded to both Mo atoms. The HOMO/LUMO gap they calculated is large (1.5 1 eV), indicating the presence of a strong 6 bond.32 This is consistent with the fact that the MoMo distance in this complex is particularly short (2.09 A).33 A comparison of the calculations on the three complexes leads to the conclusion that carboxylato ligands favor the formation of a strong MoMo quadruple bond, as found for [M02(02CH)~],3~ while with the more covalent TMTAA ligand the metal-metal bonding interactions compete with the metalligand one. Compound 1,which contains the two types of ligands, appears to be intermediate between these two situations. This ligand effect can be attributed to the difference in the *-donating ability of the ligands. The 6 frontier orbital of each MoL4 "monomeric" subunit affords some destabilization arising from interactionwith a x-typeligand orbital of the same local symmetry. This destabilization of the Mo 6 frontier orbital will induce a similar destabilizationof their bonding combination in the "dimer" and therefore reduce the HOMO/LUMO gap of the dinuclear complex. In addition, the mixing of the "monomeric" 6 frontier orbital with the *-type ligand lowers the metallic character of the 6 frontier orbital. The larger the mixing, the lower the overlap between two 6 "monomeric" orbitals and consequentlythe weaker the MoMo 6 bond in the "dimer". Oxygen atoms are weak *-donors; therefore carboxylatoligands will favor larger HOMO/ LUMO gaps and stronger 6 bonds. In contrast, nitrogen ligands such as TMTAA, which are better *-donors, will favor smaller HOMO/LUMO gaps and weaker 6 bonds. Finally, it is interesting to note a typical feature found in the three calculations: the 6* LUMO is separated from the other vacant orbitals by a gap larger than 1 eV. Consequently, in our calculations as in those of Norman et al.,31this level is situated in the middle of a large gap separating the HOMO from the second lowest vacant orbital. (31) Norman, J. G., Jr.; Kolari, H. J.; Gray, H. 8.; Trogler, W. C. Inorg. Chem. 1977, 16, 987. (32) An ab initio study on [Mo2(02CH)4] has also k e n published: Atha, P. M.; Hillier, 1. H.; Guest, M. F. Mol. Phys. 1982. 46, 437. (33) Cotton, F. A.; Norman, J. G., Jr.; Stults, B. R.;Webb, T.R.J . Coord. Chem. 1976, 5, 217.

Kerbaol et al.

718 Inorganic Chemistry, Vol. 32, No. 5, 1993 A ll 116 b21

a

I

L

Y

I

nul15all

llg 115a11

i. .

,..

L

I

1

Y

Ulllall

=&

\

\

i,

i

2

t

L

24

x

I

L

Y

1

0

I

-u

POTENTIAL

I

I

I

T

._+

.-

t,

1

f x

-1.0

U

]

V/Fc

Figure 7. Cyclic voltammetry of (a) [Mo2(02CCH3)4](h), (b) [ { M e ) ~ ] in T H F (TMTAA))2] (3), and (c) [ M o ~ ( T M T A A ) ( O ~ C C H ~(la), containing0.2 M BQNPF~. Key: Cca. lW3 M; platinum diskelectrode; scan rate 100 mV/s; reference Fc+/Fc.

I

L

1

1

-11

6*11rbll

6Vqi

t"

Figwe 6. Contour map of the metallic M O s in the X I and yz planes (contour values *0.01, h0.02, hO.04, hO.08, h0.16): (A) a(MoMo), w(MoMo), w,(MoMo); (B) d(MoMo), d*(MoMo).

.-

A

A comparativepicture of the electrochemical behavior of the three types of complexes, [Mo2(TMTAA)(02CRM (11, [M02(02CRhI (21, and [Wo(TMTAA)M (3) is presented in Figure 7, in the case of R = CHs, and the electrochemical data are presented in Table VII. [Mo2(TMTM)(O&R)J (1). In PhCN, the pivalatederivative l b gave the same response as the acetate la; the study was developed with the acetate derivative due to the low stability of

the pivalate compound. As shown on the typical voltammogram in Figure 7, curve c, both compounds, with R = CH3 or C(CH&, display in the electrochemical range of PhCN, THF, and CH2C12, two reversible redox proceases followed by a totally irreversible peak (not shown in the figure). These first two processes have cyclicvoltammetry (CV) characteristicsindicativeof one electron diffusion-controlled reversible exchanges for scan rate u, 0.01 < u < 1 V/s (Le. AEpz 60 mV, ip/ips 2 1, and i,/u1/2 is constant); they correspond to two oxidation wave8 with equal height at a rotating disk electrode (RDE). In the reduction range, two partially reversible systems can be seen at a potential very close to the THF reduction limit. The two cathodic peaks of these systems have exactly the same height and shape as those of the two fvst oxidation processes. They display a reoxidation anodic peak which can only be distinctly seen for scan rate, u, more than 0.1 V/s and whose intensity increases with u. Owing to the very low value of the potentials of these systems, it can be deduced that the reduced forms of the complexes are very reactive and have a very short lifetime. Bulk electrolysisof a solution of [MO~(TMTAA)(O~CR)~J at a potential more positive than the first process generates, after an exchange of n = 0.9 f 0.1 electron per molecule, a solution whose CV is the same as the initial one over a few hours, reflecting the stability of the oneelectron oxidized species. Conversely, the product obtained through the second one-electron oxidation process is not stableon the electrolysis time scale. This is indicated by the degradation of the CV after bulk electrolysis. The

Inorganic Chemistry, Vol. 32, No. 5, 1993 719

[Mo~(O~CR)~(C~~H~~N~)I Table MI. Electrochemical Data for the MoMo Complexes:

(AE./mV)b

E’O/YO

redn steps complex [MOZ(~~CR)~I R CH3 R

C(CH3)3

solvent

E‘O5

( u p )

PhCN CH2Cl2 THF PhCN CH2C12 THF

E’Ol ( U p )

E’O2 ( M p )

E’O3 ( u p )

-0.01 (60) 0.06 (80) -0.10 (80) -0.02 (60) 0.02 (80) -0.14 (1 10)

E, = +1.38 (irr)

-0.01 (70) 0.12 (80) -0.12 (80) -0.02 (70) 0.12 (80)

-0.12 (80)

-3.32 (180)

-3.16 (120)

-0.73 -0.62 -0.82 -0.74 -0.66 -0.86

-2.55 (300)

-0.84 (70) -0.96 (70) -0.88 (80)

-3.10 (100)

-3.38 (140)

E, c

oxidn steps

E’O4 ( U p )

0 Obtained from cyclic voltammetry in solvents + Bu4NPF6 (0.2 M); Pt disk; 100 mV/s; Separation between the two main peaks.

coulometricvalue is typically n = 1.8 f 0.3 electrons per molecule, when the electrolysis is interrupted at a current level a tenth to a twentieth of the initial one. A very long exhaustive electrolysis leads to a total degradation of the solution. Among the new peaks okerved in the CV after electrolysis the most prominent ones correspond to pseudoreversible processes located at ca. 0.40, -0.72, and -2.22 V. These are typical of a [MoO(TMTAA)]+ derivative,l6 a species whose presence is also indicated by the appearance of its ESR spectrum (vide infra). Similar observations could also be made when electrolysiswas performed at a potential corresponding to the first oxidation over a long period of time at a current level close to the background. The source of oxygen in the formation of the oxo species is probably the residual water in the solvent although invery low concentration. Bulkelectrolyses performed at potentials more cathodic than the two reduction peaks failed to produce any result other than totally decomposed solutions. Attempts to chemically (NaHg) reduce the complexes 1 also failed. [{MO(TMTAA)}~] (3). We have previously described briefly the electrochemical behavior of [(Mo(TMTAA)}~];~ for comparison, it was reexamined here. Moreover, a misinterpretation in the assignment of the redox processes was found in the earlier work.9 As shown in Figure 7b, [{Mo(TMTAA)}2]gives rise to two fully reversible processes, at -0.84 and -0.40 V in PhCN, followed bya third more complex one. Thelatter oxidation process displays two anodic peaks associated with a large cathodic peak and two smaller cathodic peaks shown by the dashed line in the figure, one of them at a much more negative potential, ca. -0.7 V. In THF, a reduction process is observed at -2.55 V; the height of the cathodic peak is a little less than twice that of the two reversiblepeaks; the anodic peak of this system is very broad with AEp = 340 mV. At a rotating disk electrode, a solution of the neutral [{Mo(TMTAA)}2] compound shows fwo successiue oxidation waues of q u a l heights, corresponding to the first two reversible processes, and not one reduction and one oxidation waue as previously reported.9 These are followed by a more complex composite wave, corresponding to the third oxidation process, whose height q u a l s that of the sum of the first two. The same height was found for the RDE wave of the cathodic process. Bulk electrolysis performed on the complexes in PhCN, at a potential more anodicthan the first waves, gave coulometricvalues q u a l to n = 0.9 f 0.1 faraday/mol. The oxidation product of the first process has the same electrochemical characteristics as the starting material, indicating an uncomplicated oxidation process. Conversely, after oxidation at a potential more anodic than the second wave, as is observed in the case of the

E’O

(60) (80) (80)

(60) (80) (80)

-0.40 (70) -0.44 (70) -0.42 (80)

E, = +1.38 (irr)

0.44 ( 1 1O)c

E , = 0.48 0.40 ( 120)c

versus Fc+/Fc reference electrode. AEp = E, -E,.

[Mo~(TMTAA)(O~CR)~] complex, the RDE and cyclic voltammograms indicate the formation of a mixture of [(Mo(TMTAA)J212+and [Mo0(TMTAA)]+.l6 No electrolysis was performed at the potential of the third oxidation process. However, the CV is indicative of a coupled chemical reaction through the presence of several cathodic peaks associated with the anodic peak and with a decrease of those of the starting material, while peaks characteristic of [MoO(TMTAA)]+ appear when the potential sweep is interrupted during a few minutes at a potential more anodic than that process. Every attempt to generate a reduced species, at potentials more cathodic than each of the two reduction waves, failed and gave decomposition products. [M02(02CR)4] (2). For the purpose of comparison, the electrochemical behavior of the tetracarboxylato complexes [M02(02CR)4] with R = CH3 (Figure 7a) or C(CH3)3, although previously briefly described,12was reexamined. Contrary to what was previously anticipated, these derivatives are soluble enough to allow electrochemical and ESR studies; the pivalate complex is more soluble but is less stable in solution. These derivatives give rise to an oxidation process whose cyclic voltammetry (CV) characteristics are indicativeof a oneelectron diffusion-controlled reversible exchange; this first process is followed by a totally irreversible wave close to the solvent discharge. Down to -3.5 V vs the Fc+/Fc electrodein THF, no reduction system is observed. Exhaustive electrolysis of the solutions, at a potential correspondingto the first oxidation wave, indicated coulometric values of n = 1.O f 0.1 faraday/mol and generated solutions of the oxidized species stable over a few hours, sufficient for ESR characterization. ESR. A set of spectra recorded at room and low temperature (130 K)in PhCN is presented in Figure 8 for the three kinds of compounds with R = CH3, oxidized at a potential corresponding to the first oxidation wave. Data are reported in Table VIII. The ESR spectra of the reduced solutions of [Mo2(TMTAA)(02CR)2] and [{Mo(TMTAA)Jz]were noninterpretable, with very low intensity. [MO~(TMTM)(O~CR (1). ) ~ ]Theisotropic room temperature spectra of the two carboxylato derivatives, h , b , after oxidation at a potential corresponding to the first wave, show the usual pattern of three components of one, six, and eleven lines, as expected for a delocalized electron on two equivalent molybdenum nuclei (92-94s96-9*JwMo, I = 0, natural abundance ca. 75%; 95-97Mo, I = 5/2, natural abundance ca. 25%).*1J21J3*The lowtemperature spectra show very poor resolution, and it was not possible to determine whether they correspond to a delocalized

Kerbaol et al.

720 Inorganic Chemistry, Vol. 32, No.5, 1993

t:g=ZDoo

Fi-8. ESR spectra of the molybdenum complexes in benzonitrile after electrolyses at potentials more anodic than the first oxidation waves: Upper part, room temperature (298 K); lower part, low temperature (130 K). Key: (a) [Mo2(02CCH3)4] (28); (b) [MO~(TMTAA)(O~CCH&] (la); (c) [{Mo(TMTAA))21 (3).

Table VIII. ESR Data for the Oxidized Forms of the MoMo Complexes" complex a s 0 Aim B~I [Mo2(02CCHd41+ 1.932 28 1.933 [Moz(02CC(CHddrl+ 1.933 26 [MO~(TMTAA)(O~CCH~)~]+ 1.941 22 1.958' [MO~(TMTAA)(O~CC(CH~)~)~]+ 1.942 22 1.955' [{Mo(TMTAA))21+ 1.954

g,

AII

1.933 41 1.934b 40 1.936' 1.937' 1.956'

M. A in gauss throughout a In PhCN, 0.2 M Bu4NPF6; C ca. this paper, 1 G = 10-4 T; 298 K for the room-temperature spectra and 130 K for the low-temperature spectra. Central line. Most intense low-field line.

mixed-valence form in a frozen solution. These spectra appear to be composed of a central line and a less intense left-hand line, tentatively referred respectively as g, and gll in Table VIII; moderately intense lines on the right-hand side of the spectra could however be assigned to the third component of a rhombic spectrum. This shape seems to indicate a rhombic geometry is more likely than the axial arrangement observed for the tetracarboxylato derivatives (vide infra). Contrary to expectation, the solutions oxidized at a potential more anodic than the second reversibleprocess are not ESR silent. During the oxidation the intensity of the initial signal of [Mo~(TMTAA)(O~CR)~]+ decreases, while a new signalappears and increases in intensity. The characteristics of this signal at room temperature, gw = 1.958 and Ai, = 35 G, as well as the low-temperature spectrum, are those of the previously described [MoO(TMTAA)]+ derivative.I6 The latter signal becomes the only one observablewhen the electrolysisis continued exhaustively. This corroborates the results of the electrochemistry. No other signal could be observed at any time during the electrolysis, indicating that the oxidized form, [Mo2(TMTAA)(02CR)2I2+, detected by RDE voltammetry, is ESR silent, as expected. [(Mo("TAA)}2] (3). The product of the first one-electron oxidation process displays ESR spectra which are very broad both at room and low temperature, as shown in Figure 8c; they do not show any resolved coupling. This is probably due to the very large value of the line width (central line line width ca. 53 G), precluding any resolution of the satellite lines which should have coupling constants in the range of 20-60 G. In the roomtemperature spectrum, deformations on each side of the central line are ascribable to the six-line set, while those corresponding

to the putative eleven-line set for a delocalized mixed-valence [Mo,~~%Mo,~~']+ derivative, with I M = ~ I M =~5/2,~ cannot be distinguished from the base line. Although the low-temperature spectrum could not be resolved (Figure 8c) its shape is consistent with an axial geometry; cf. the tetracarboxylato derivatives. As for the other TMTAA derivatives I, the solution resulting from the second oxidation is not ESR silent but only displays the typical ESR signal of [MoO(TMTAA)]+, both at room and low temperature (vide supra). ~ 0 4 0 2 C R ) , ](2). The spectra obtained for both compounds, with R = CH3 and R = C(CH3)3,after oxidation at a potential corresponding to the first wave, are very similar, as previously described.l'J2bJ7bJ4 are also similar to that of the derivative with R = c3H.1.'~They all display symmetric character with regard to the central line and are interpretcdasanaxialspectrumwithgvaluesoftheperpendicular and parallel components nearly equal. This interpretation reflects the axial geometry of the complexes. As expected for a delocalized mixed-valence [Mo").~MoI~~]+ compound, the three sets of 1,6, and 11 linesare observed on the parallel component of the spectra. In the case of the present compounds, the isotropic roomtemperature spectra could also be obtained, and in these the three sets of lines are also observable and the eleven-line set is clearly distinguishable, as shown in Figure 8a. Comments on the Electrochemid Reactivity of tbe [Md Mol Quadruple Bond. The present results exemplify the electrochemical reactivity of the dimolybdenum quadruple bond, which can be summarized by the following scheme:

[MoOLI+

+ Degndation products Degradation producu

The [Mo"~Mo"] core can be oxidized in two reversible oneelectron steps. The first step leads to stable mixed-valence [Mo"'.5MoI1']+ class 111 compounds.11 in which the odd electron is delocalized over the two molybdenum nuclei. The ~~~

~~

(34) Zhuang, Q.;Fukuoka, A.; Fujimoto, T.; Tanaka, K.; Ichikawa, M. J .

Chem. Soc., Chem. Commun. 1991, 145.

Inorganic Chemistry, Vol. 32, No. 5, 1993 721

t

derivatives.gJ1-13Thestrongest quadruplebond in [M@(02CR)4], explained by the theoretical analysis, is illustrated by its redox ----behavior. This compound is the most difficult to oxidize to [Mo"~Mo"']+, and this is the only redox process observed in the electrochemical window of common solvents. The electrondonating effect of the TMTAA ligand weakens the MoMo quadruple bond, and accordingly, the derivatives are easier to oxidize and to reduce; this effect is more pronounced when two TMTAA ligands are present (Table VII). However, it is difficult to make a comparison, because the coordination mode of the TMTAA ligand is different in the two complexes, 1 and 3, as shown in Figure 2. The difference in the coordination mode of the TMTAA ligand is also illustrated by the ESR spectra. The low-temperature spectra reflect a same kind of symmetry, axial or isotropic, for the compounds [Moz(OZCR)~]and [(Mo(TMTAA)I2], while the spectrum probably corresponds to orthorhombic character in [Mo~(TMTAA)(O~CR)Z], probably Figure 9. Potential-stick diagram showing the effect of the coordination owing to a lower symmetry. This is in good agreement with the modeon thevalueoftheredoxproctssesforthe threecomplexes [M02(02schematical structures shown in Figure 2. CCHhI (%I), [MOZ(TMTAA)(OZCCH~ZI (14, and [IMO(TMTAA)IZI As far as the HOMO-LUMO gap is concerned,an observation (3). The difference bctwccn the first oxidation (Eox')and first reduction similar to that obtained by the theoretical analysis can be made (Erdl)potentials, represented by AE,is an estimation of the HOMOLUMO gap. from the electrochemicalmeasurements. The difference between the potential of the first oxidation process and the first reduction reaction is generally considered to provide an evaluation of the second oxidation step represents the transient formation of a HOMO-LUMO gap, at least comparatively in a series. The [ M O ~ ~ ' ~ M O ' 'derivative, ']~+ which decomposes very quickly, values obtained in the present case, from Table VII, can be owing to the weakening of the bond order. Reduction steps are also evident at ca. -3.1 and -3.3 V for [Mo~(TMTAA)(O~CR)Z] estimated as greater than 3.38 V in the case of [ M O ~ ( O ~ C R ) ~ ] and equal to 2.30 and 1.67 V, respectively, for the derivatives (1) and -2.5 V for [(Mo(TMTAA)JJ (3) and are ascribed [Moz(TMTAA)(OZCR)Z]and [{Mo(TMTAA)Jz]. This is ilto a two- or one-step reduction of [Mo"~Mo"] into lustrated by the diagram of Figure 9, which displays well the [Mo'FMo'l2-. In this case also, the decrease of the bond order effect of the coordination mode on the electrochemical levels. is probably the origin of the instability of the reduced species. Their ordering compares quite well with that obtained by the This precluded their characterization by any means and especially theoretical analysis and are in full agreement with the observed by ESR for the [ M o " ~ M o ' ] -derivatives; this type of complex modulation of the [ M d M o ] bond strength by the type of has been scar~elydescribed and WBS difficult to ~haracterize.'3b9C*435 coordination. The application of these observations to the The assignation of the site of the reduction, in the present case, reactivity of the [ M d M o ] unit toward small molecules, which is attested by the metallic character of the LUMO '6 orbital recently found a renewal of intere~t,3~J6J~ is currently under involved, demonstrated in the theoretical analysis. Moreover, investigation. the differencesbetween theoxidation and f m t reduction potentials calculated in the case of other Moz c o m p l e ~ e s Iare ~ ~ in ~ -the ~ AcLWwIedgmemt. The CNRS (URA 322, URA 254, and URA same order of magnitude. In the previous work? the misassign804) is acknowledged for support of this work. We thank Dr. A. ment of the first reduction process of [{Mo(TMTAA)J2](3) to Le Beuze for helpful comments and Pr. S. A. R. Knox for a the reversible step located at ca. 4 . 8 2 V in THF was probably careful reading of the manuscript. due to the presence of the oxidation product of this reaction by suppkmenCuy MateM AvrihMe: For compound [Mc~(CaHaN4)(02action of dioxygen. This was also observed here in some cases, CCH&], tables giving crystal data and details of the structure but the effect of dioxygen on this family of compounds has not determination, positional parameters, bond distances, bond angles, been scrutinized yet. anisotropic thermal parameters for non-hydrogen atoms, equations of The present study shows that theoccurrenceand characteristics selected least-squares planes and interplanar angles, and torsional angles of these different redox steps are strictly under the control of the (1 1 pages). Ordering information is given on any current masthead page. coordination mode of the [ M d M o ] moiety. These results are in keeping with those obtained for the previously described (36) Kerby, M.C.; Eichhorn, B. W.; Doviken, L.; Vollhardt, K. P.C. Inorg.

lEJxL

IEo V/Fc

I-1-

- -

(35) Baxendale, J. H.; Garner, C. D.; Senior, R.G.; Sharp, P. J . Am. Chem. Soc. 19'16, 98,637.

Chcm. 1991, 30, 156. (37) Eichhorn, B. W.; Kerby, M.C.; Haushalter, R.C.; Vollhardt, K. P.C. Inorg. Chem. 1990, 29, 723.